Probabilistic models for CRISPR spacer content evolution.

Area of Science:

• Microbial genomics
• Molecular evolution
• Bioinformatics

Background:

• The CRISPR/Cas system provides adaptive immunity in bacteria and archaea through heritable spacer sequences.
• Spacer content evolves rapidly and dynamically, differing from standard nucleotide evolution models.

Purpose of the Study:

• To present novel probabilistic models for quantifying CRISPR/Cas spacer content evolution.
• To account for diverse insertion and deletion processes affecting spacer arrays.

Main Methods:

• Developed probabilistic models incorporating constrained or unconstrained insertions and single-spacer or fragment deletions.
• Estimated model parameters using maximum likelihood and explicit ancestor enumeration.
• Simulated data to validate model performance and parameter estimation accuracy.

Main Results:

• Models accurately estimate evolutionary rates and divergence times between bacterial strains.
• Spacer overlap limits the estimable evolutionary distance, establishing an upper bound.
• Phylogenetic analysis of *Yersinia pestis* data revealed congruent results and captured diversity variations.
• CRISPR array phylogenies and Cas gene phylogenies indicated different evolutionary time scales.

Conclusions:

• The developed models effectively quantify spacer content evolution and visualize strain similarities in phylogenies.
• Enables comparative analysis of CRISPR arrays and integration with nucleotide substitution rate data.